Please use this identifier to cite or link to this item:
|Title:||A Review of Image Quality Assessment Algorithm to Overcome Problem of Unnatural Contrast Enhancement||Authors:||Nur Halilah
Chen Soong Der
|Issue Date:||2015||Conference:||The 3rd National Graduate Conference (NatGrad2015), Universiti Tenaga Nasional, Putrajaya Campus, 8-9 April 2015||Abstract:||Image contrast enhancement is designed to increase the visibility of image details. The following problems may occur after contrast enhancement: amplification of noise artifacts, saturation-loss of details, excessive brightness change and unnatural enhancement. The objective of this paper is to survey the existing Image Quality Assessment Algorithms (IQAs) to overcome the problem of unnatural contrast enhancement. The aim of IQA is to rate the quality of the image in a way consistent to human visual perception. The rating is useful as feedback to image processing algorithm to fine tune parameters for optimal result. The survey shows that most IQAs are fidelitybased which are not suitable for evaluating of the contrast enhanced image. The majority of non-fidelity-based IQAs are related to sharpness and contrast. There are only a small number of IQAs designed to address the problem caused by contrast enhancement. There are two IQAs designed to address the problem of unnatural contrast enhancement – Lightness Order Error (LOE) and Structure Measure Operator (SMO) but they are found not giving a rating consistent to human visual perception. Statistical Naturalness Measure (SNM) is designed to evaluate the quality of Tone Mapped images. It is found that one of statistical model the IQA which is related to the contrast is useful in detecting the problem unnatural contrast enhancement. However, the problem of inconsistent ratings for same image content but different spatial resolution needs to be addressed.||URI:||A Review of Image Quality Assessment Algorithm to Overcome Problem of Unnatural Contrast Enhancement
|Appears in Collections:||COGS Scholarly Publication|
Show full item record
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.